So, here's the data I have:
Microarrays showing DEGs from cell populations treated with 5-aza and untreated cell lines. Bisulfite seq data for the same cell lines with the same treatments.
My question is: How can I compare/associate methylation +/- 200 bp from start of transcription in DEGs with changes in expression on a genome wide scale without having to look (at least initially) at individual genes at the sequence level? Are there any good programs/approaches for analyzing this type of association on a genome wide scale?
Ideally, I'd like to identify genes that exhibit both >50% reduction in methylation about the start of transcription, and a greater than 3-fold increase in expression. Doing this for the microarray is fairly straightforward, but pulling all the corresponding sequences individually from the bisulfite data is a daunting task involving thousands of loci. There must be a better way. Is there a better way?
I think there should be a way to annotate the bisulfite seq data. you can have a look at HOMER bioinformatic tool.
You can first analyze both data sets independently. then you must use one common identifier to cross both data sets, after applying your filters.
For crossing the two tables I recommend you KNIME. Gene name can be your identifier.
Best wishes,
Adriana
Hi Carl,
Have you tried loading your microarray data and methylation data together in IGV (Integrative Genomics Viewer)? You could then apply filters to visualize relevant regions of the genome. Note this would be an exploratory approach.
On another subject I just made a small video to explain our website and would appreciate your feedback. Please share it if you think it's relevant.
Best,
Alain
https://vimeo.com/125323142
Alain, thank you very much for sharing your video. It remembered me myself few years ago. I will try Insilico DB from now on. Best wishes.
How long did you treat your cells with 5-aza before analyzing gene expression?
I have done the 5-Azadeoxycytidin (AZA) treatment at 50 nM concentration for 48 hours on adherent cells for further Expression analysis and MS-PCR.
I see a potential problem here. 5-aza is a cytidine analog that inhibits methyltransferase, thereby preventing methylation of NEWLY-SYNTHESIZED DNA. It does not strip methylcytosines off existing DNA molecules. Forty-eight hours corresponds to at most 2 doubling times for your cell population (probably less due to toxicity of the agent). Nevertheless, if we assume 24 h doubling, at 48 h ~1/4 of your cells will have normal methylation, ~ 1/2 will have ~50% depletion (assuming complete inhibition by 5-aza) and 1/4 will be methylation deficient. Given that 5-aza does not completely inhibit methylation, in practice these numbers will be greatly skewed toward normal methylation. Thus, any gene expression differences you measure after 48 h of treatment are unlikely to be due to inhibition of methylation.
I was changing the medium after 24 hours. You are right with your statement. Would you recommend to keep the cells for longer time replenishing AZA treatment every two days and passaging the cells for some weeks? and then test expression and DNA methylation profiles?
For the purpose of my experiment it was fine to do it like that. I wanted to compare with another type of treatment and other approaches.
Yes, definitely. Continue to grow the cells for as long as is experimentally and scientifically feasible. Make sure you assess cell growth, so you can estimate the number of doublings. Most importantly, however, when you do the expression analysis, you should simultaneously analyze the same cell populations using bisulfite sequencing to ensure that the desired alterations in methylation did in fact occur. Otherwise, you might find that a subclass of 5-aza resistant cells are what continue to grow in the plate, while the cells that are actually inhibited could be arrested.
DNA Methylation Data Analysis Workshop
How to use bisulfite-treated sequencing to study DNA methylation
When?
12-15 November 2019
Where?
Berlin, Germany
Link?
https://www.ecseq.com/workshops/workshop_2019-07-NGS-DNA-Methylation-Data-Analysis
- Badges
- Science method